WLCG

WLCG
Name	WLCG
Caption	Worldwide LHC Computing Grid conceptual diagram
Type	Distributed computing infrastructure
Founded	2001
Headquarters	CERN
Area served	Worldwide

WLCG WLCG is a global distributed computing infrastructure created to provide the large-scale data processing and storage required by high-energy physics experiments at the Large Hadron Collider, coordinated from CERN. It federates computing and storage resources across national laboratories, regional data centres, and university clusters to support experiments such as ATLAS, CMS, ALICE, and LHCb. The project connects major institutions like Fermilab, DESY, INFN, STFC, and TRIUMF to enable worldwide access to petabyte- and exabyte-scale datasets produced by collider operations and physics analyses.

Overview

The infrastructure combines grid middleware, networking, and site operations to present an integrated service for experiment collaborations such as ATLAS Collaboration, CMS Collaboration, ALICE Collaboration, and LHCb Collaboration. It relies on tiered resource models involving CERN as a central hub and regional centres in continents represented by organisations like National Research Council (Italy), CNRS, Deutsches Elektronen-Synchrotron, and Science and Technology Facilities Council. Key technologies and projects integrated include gLite, HTCondor, ARC (Advanced Resource Connector), XFEL, OpenStack, and Ceph, enabling workflows developed in environments such as ROOT (framework), Geant4, Gaudi (software), and PAT (CMS).

History and Development

The initiative emerged after design studies and community workshops involving stakeholders from CERN, Fermilab, Brookhaven National Laboratory, and national agencies like DOE and EU funding programmes such as Framework Programmes for Research and Technological Development. Early milestones included prototype demonstrations at events like the ACM conferences and production deployments supporting first collisions in the Large Hadron Collider commissioning phases. Development cycles incorporated collaborations with projects including EGEE, Enabling Grids for E-sciencE, and later integration with cloud initiatives from Amazon Web Services, Google Cloud Platform, and research clouds coordinated through EUDAT and GEANT.

Architecture and Components

The topology follows a tiered model with central archival and distribution roles played by CERN and regional hubs operated by organisations such as GridPP, NorduGrid, OSG (Open Science Grid), and NDGF. Components include storage fabrics built on technologies like dCache, StoRM, and CASTOR; compute layers using batch systems like PBS Professional, SLURM, and HTCondor; identity and access services integrating Kerberos, LDAP, and federations such as eduGAIN. Data management integrates catalogues and transfer services exemplified by FTS, Rucio, and PhEDEx, while monitoring and accounting leverage tools from projects like Nagios, Grafana, and ELK Stack adopted by centres including KIT, CEA, Rutherford Appleton Laboratory, and CNAF.

Operations and Governance

Operational coordination is organized through programme boards, technical boards, and operations teams involving representatives from experiments and funding agencies including European Commission, National Science Foundation, and national ministries such as Ministry of Education and Research (Germany). Formal governance includes service-level agreements between CERN and regional providers along with formal reviews by panels comprising experts from IHEP, KEK, TRIUMF, and academic groups from universities like University of Oxford, MIT, University of Tokyo, and University of Melbourne. Day-to-day activities are executed by site administrators, grid operators, and experiment run coordinators who interface with networks such as GÉANT and undersea cable consortia connecting nodes in North America, Europe, Asia, and Australia.

Performance and Usage

The infrastructure sustained processing for major discovery campaigns, including the analysis that led to the observation announced by teams within ATLAS and CMS related to the particle observed at CERN in 2012, enabling large-scale Monte Carlo production and reprocessing with tools from ROOT (framework) and Geant4. Performance metrics track throughput, successful job completions, and transfer rates measured in gigabits and petabytes across backbones serviced by providers like Lumen Technologies and research networks coordinated through GÉANT. Usage spans thousands of physicists from institutions such as Imperial College London, University of California, Berkeley, École Polytechnique, and Tsinghua University performing analyses, producing publications in journals such as Physical Review Letters and Journal of High Energy Physics.

Collaborations and Impact

The project catalysed collaborations among national laboratories including Brookhaven National Laboratory, Lawrence Berkeley National Laboratory, Los Alamos National Laboratory, and international consortia such as Open Science Grid and EGI. Its technologies and practices influenced data-intensive projects in domains represented by institutions like CERN Openlab, Human Genome Project, and observatories like Square Kilometre Array and European Southern Observatory. Training and outreach involve summer schools and workshops hosted by CERN, universities, and organisations such as ITU and IETF, while policy and open-data initiatives intersect with publishers and funders including Nature (journal), Science (journal), and the European Research Council.

Category:Distributed computing Category:High-energy physics